Resumen de Comparison of experimental designs for normal and gamma distributions

Víctor Manuel Casero Alonso, Sergio Pozuelo Campos, Mariano Amo Salas

• The aim of this work is to show the e ect of misspeci cation in the probability distribution on optimal design. A generalized Fisher information matrix is obtained using the elemental information matrix, which includes information on the probability distribution of the response variable. Relevant theoretical results were obtained, for di erent regression models, comparing heteroscedastic gamma and normal distributions. Finally a practical case which considers a 4- parameter Hill dose-response model is used.